Never seen this before. Seems to be the response from all the AI. Is there a service outage or is it a problem on my end?
I have the same error. I don’t think it’s a problem on your end.
I received the same error.
I also get this error and asked deepseek about it. Got the following answer:
The error message no deepseek v4 endpoints found, deepseek/flash indicates that the system (likely a third-party tool named “Autoblow”) is trying to connect to the DeepSeek API using an incorrect endpoint version. Here’s what you need to know:
Key Issues:
-
DeepSeek does not have a “v4” API – The current stable API version is v1 (e.g.,
https://api.deepseek.com/v1).
Usingv4will always fail. -
“flash” might refer to a model name (like
deepseek-chatordeepseek-reasoner) but there is no official “flash” endpoint in DeepSeek’s API. -
The tool (“Autoblow”) may be outdated or misconfigured. Even if it claims to use DeepSeek, the error shows it’s referencing a non-existent API version.
How to Fix:
-
Check the API base URL in the tool’s settings. It should be:
https://api.deepseek.com(withoutv4).
Add/v1for the chat completion endpoint:https://api.deepseek.com/v1/chat/completions. -
Use the correct model name:
deepseek-chat(ordeepseek-reasonerfor reasoning).
Remove any reference to “flash”. -
Update the tool – If “Autoblow” is a third-party client, check for updates or contact its developer. The error suggests it’s using an outdated API specification.
-
Verify your API key – Ensure the key is valid and has access to the intended models.
If the problem persists, you might need to switch to a different client that properly supports DeepSeek’s actual API endpoints.
Sorry guys, the girls (and some men) were off for 12 hours. We had some issues updating our LLM model. The LLM model is updated now and working. They stopped serving the old model and we did not catch the need to update to the new one.
1 Like